Automated Correction of Bridge Slab-Bending Affect in Terrestrial Laser Scanning

Abstract

A 3D measurement, such as a terrestrial laser scanning, is applied for an advanced infrastructure management and Building Information Modeling (BIM). Although a terrestrial laser scanning can acquire massive point cloud data for the BIM, 3D measurement using high precision LiDAR is affected by slab-vending with active loading, such as vehicle movements on a bridge. Thus, stripy noises occur in acquired point clouds. Therefore, we proposed three methodologies, such as multiple data subtraction, plane estimation with Least Median of Squares (LMeds), and noise pattern estimation, to remove the stripy noises for precise 3D bridge modeling. Our three methodologies were verified using terrestrial LiDAR data taken under a road bridge. Through our experiments, we confirmed that our algorithms can automatically cancel the vending affect of bridge in laser scanning works.